Groche is a smart grocery assistant inspired by the common problem of buying food without a clear plan, which often leads to waste or missing ingredients. We built it as a full-stack project using HTML/CSS/JavaScript for the frontend and Python with Flask for the backend, where users can input recipes with quantities and units, and the system matches them against a live shopping cart. We also designed it to integrate with YOLOv8 for real-time item detection, adding a layer that converts raw detections into structured data. Through this project, We learned how to connect frontend and backend systems via APIs, design consistent data models, and handle real-world issues like user input errors and unit mismatches. One of the biggest challenges was restructuring the system to include units and ensuring everything stayed consistent across components, as well as debugging API communication and UI behavior.

Built With

Share this project:

Updates